OBJECTIVE To compare Humphrey Matrix (Matrix) with Octopus perimetry for assessing visual field (VF) indices, size, severity, and location in the glaucomatous VF defect. METHODS Matrix (30-2 threshold program) and Octopus 1-2-3 perimetry (G1X program) were conducted on 60 eyes of 60 patients with primary open-angle glaucoma. The VF indices derived by Matrix (mean deviation, pattern standard deviation), and Octopus (mean defect, square root of loss variance, corrected loss variance) were compared, as were the size (percentage of points depressed < 5% or < 1%) and the severity (score and Hodapp-Parrish-Anderson grade) of the VF defect. RESULTS Matrix and Octopus results agreed well overall for VF indices, size, severity, and locations of the VF defect. The mean deviation and pattern standard deviation in Matrix were significantly higher than the mean defect, square root of loss variance and square root of corrected loss variance in Octopus. On the total deviation plot, the size of the VF defect in the superior and nasal fields reported by Matrix was significantly higher than by Octopus. The VF defect score by Matrix was significantly higher than that by Octopus in the superior, nasal, and whole fields. More eyes had a worse grade by Matrix than by Octopus. CONCLUSIONS Matrix compared well with Octopus in detecting glaucomatous visual field defects. However, the defects reported by Matrix were larger and more severe than with Octopus.